R Workshop

Module 6: “To Infinity and Beyond” (2)

2018-04-18
Bobae Kang
(Bobae.Kang@illinois.gov)

• We cannot know everything.
• In fact, no one knows everything!
• “Someone has already done it.”

  Source: AZ Quotes

Source: Wikimedia Commons

• First, try yourself. If you cannot solve it, then go online.

15 min rule: when stuck, you HAVE to try on your own for 15 min; after 15 min, you HAVE to ask for help.- Brain AMA pic.twitter.com/MS7FnjXoGH

— Rachel Thomas (@math_rachel) August 14, 2016

• We all make typos. Check for typos!
  • RStudio does not check for typos automatically
  • But we can refer to an error caused by the typos
• Check if a package is loaded before using its functions
  • Error in some_function() : could not find function "some_function"

# these are equivalent
?some_function
help(some_function)

• Looking ino the documentation is often the best way to understand what a function is and how to use it.
• Using the ? followed by the function name or help() brings out the documentation if available

• When an error is thrown, it comes with an error message
• Error messages often have rich information about what went wrong and where it went wrong
• If we are working with custom functions we defined, RStudio's debugging tools can help us to spot the source of an error in the script and debug it
  • See this article on debugging with RStudio
  • See this video by RStudio on introduction to debugging

Source: Google.com

• Be specific and succint
  • The search term should be a set of keywords
  • Using the relevant error message as a serach term can help

Source: R Project

• The Comprehensive R Archive Network (CRAN)
  • Manuals page
  • Task Views page
  • Packages page

• CRAN offers the following manuals:
  • “An Introduction to R”
  • “R Data Import/Export”
  • “R Installation and Administration”
  • “Writing R Extensions”
  • “The R language definition” (draft)
  • “R Internals”

• A Task View offers a brief introduction to a particular topic and an annotated list of relevant R packages
• CRAN has tasks views on a selection of topics, including:
  • Machine learning & Statistical Learning
  • Multivariate Statistics
  • Official Statistics & Survey Methodology
  • Statistics for the Social Sciences
  • Survival Analysis
  • Time Series Analysis

• Each contributed package that is listed on CRAN has a page
• A reference manual and vignettes can be found on the CRAN package page
• To directly get to the package page, try on your broswer:
  • “https://cran.project.org/package=[package-name]”
  • Replace [package-name] with any existing package name

• Packages have reference manuals that contain documentation for all its contents (i.e. functions and datasets)
  • Basically, it is a collection of help() documentations in a pdf format
• Reference manual can also be found by googling
  • Try “package-name pdf” as your search term

• Packages often have vignettes to introduce its contents
  • Some vignettes can be accessed via vignette("package") on console
  • Other vignettes are found on the pacakge page on CRAN
• Unfortunately, not all packages have vignettes.

• https://www.rstudio.com/
  • Cheet sheets page
  • Webinar & videos page

• Currently, 13 RStudio cheat sheets are available, including:
  • “Data Transformation with dplyr”
  • “Data Import”
  • “Data Visualization with ggplot2”
  • “Date and times with lubridate”
  • “Work with strings with stringr”
• Currently, there are 15 user-made cheat sheats as well
• Some cheat sheats can also be found in RStudio IDE menu
  • “Help > Cheatsheets”

• https://www.tidyverse.org/

• Articles

• https://shiny.rstudio.com/tutorial/
  • Video & wrttien tutorial page
  • Articles page
  • Reference page

• http://www.htmlwidgets.org/

• https://www.r-bloggers.com/

• Wickham H. and Grolemund G., R for Data Science
• Wickham H., Advanced R
• Wickham H., R Packages
• U of Cincinnati, UC Business Analytics R Programming Guide
• Wikibooks, R Programming

Source: R for Data Science

• Yau, C. R Tutorial
• Kabacoff, R. Quick-R

Source: GitHub

“GitHub is a development platform inspired by the way you work. From open source to business, you can host and review code, manage projects, and build software alongside millions of other developers.” - GitHub.com

• Most R packages are available as GitHub repositories, which can be “cloned” and downloaded if wanted.
• Many R package authors offer brief explanations and even quick tutorials for their packages on the GitHub repositories.

Source: worldview.stanford.edu

• DataCamp
• Coursera
• edX
• Udacity

• Requires registration and log-in
• Some free courses are available, but most are paid courses with one free chapter
• Cost is $25/month with the annual plan or $29/month
  • Onces subscribed, all courses become available
• Offers 70+ courses on R
  • Each course is short (~4 hours) and focused on a specific topic
  • Ranging from basic to intermediate level

• Requires registration and log-in
• You can “audit” the course for free
  • Certification for course completion and online support become available onces you purchase the course
• Notable contents
  • Data Science Specialization (10 courses)
  • Statistics with R Specialization (5 courses)

Source: